Typically, multiple displays are synchronized based on specific starting times for each of the displays. Objects that span the multiple displays are deterministic so as to appear to be synchronized but the objects are not capable of handling user interaction. Sometimes animated objects may appear synchronized by synchronizing time and animating the objects at pre-determined (still deterministic) times over specific intervals.
Another approach is to utilize a single computing host to control multiple displays. In this situation, the objects can be interactive and not deterministic but a single host is controlling the objects on the displays. This is similar to extending a user's desktop to two or more displays and objects presented on the display can be moved from display to display through user interaction.
However, this latter approach is inefficient for many reasons. The host can be loaded with other processing, which affects the interaction and animation of the objects adversely. Furthermore, most displays and devices with displays include computing capabilities, such that taking this latter approach leaves processing capabilities on one of the displays underutilized.
Therefore, what is needed is an approach where two or more devices can cooperate to achieve interactive animations of objects presented on the displays of the devices.